Relevance Engineering for Local SEO

AI Overviews are here, and they’ve fundamentally altered the nature of local search.

Traditional local SEO strategies are already obsolete as Google has moved on to large-scale semantic synthesis, which can be influenced by a new practice known as relevance engineering.

99% of the local SEO industry remains fixated on keyword density, link graphs, and other relics of the model that kept them employed and profitable for multiple decades.

The problem is that none of these concepts even slightly represent how Google currently selects, evaluates, and displays local content.

Industry expert Mike King of iPullRank has been one of the few who have publicly laid out this new reality:

“We need to reinvent ourselves. And that reinvention I’m calling relevance engineering… because the SEO frame is so limiting.” – Mike King, Search Engine Land Interview, May 2025

At DataPins, we are leading the charge of this reinvention and have been before Google even implemented AI Overviews into search.

Why? The writing has been on the wall for some time. Google, as the SEO community perceives it, is completely gone – even while they cling to its reputation.

Google wants to be more like ChatGPT or even Facebook—an app in which users sign in and get quick results for everyday problems (like finding a local plumber, roofer, or HVAC company).

DataPins is a Local SEO software platform enabling local contractors and home service businesses to generate job-specific content signals based on authentic locations and services.

To clarify, DataPins is a relevance engineering engine engineered explicitly for how Google’s AI processes information today.

Relevance Engineering

The Query-Fan Out Optimization Engine

Unless you’re highly tapped into Google advancements, you probably don’t know about the idea of the query-fan-out.

The query fan-out is when Google’s AI Overviews drills down the user’s search query into multiple topically related sub-queries before gathering fragments from multiple sources to synthesize a thorough answer.

Here is how Mike King describes it:

“They have this idea that they call query fan-out… they’re handing the query off to Gemini 2.5 Pro… returning a bunch of queries and different data points… performing all these searches in the background, pulling chunks from pages, and feeding that back to Gemini to generate the response.” – Mike King

For local service area businesses (SABs), this means your website doesn’t have to rank #1 for the original search query to be included within the AI Overview. 

Instead, it merely needs to be related to one or more of the hidden fan-out queries that Google generates privately.

Relevance Engineering Local Example

Why DataPins Was Engineered for Fan-Out, Not Just Search

DataPins developers understood from day one that local search is a relic compared to modern consumer behavior.

Thinking of everything as pages and keywords limits your ability to reach homeowners in your area.

That’s why DataPins created the “pin,” which generates the following signals:

  • Jobsite Records: A citation for your completed job (with images, captions, and check-in schema)
  • Structured Data: Structured data for search engines to index location, service type, timestamp, brand, customer review, and more
  • Brand Context: Authentic representations of your brand and its services
  • Longtail Keywords: The same fan-out queries that SEO fails to optimize for are generated organically through AI-structured job descriptions

DataPins is the only current local SEO tool that generates semantic entry points, each of which may be relevant to thousands of fan-out subqueries.

This contrasts with conventional SEO software aimed at reverse-engineering “keywords,” which limit brand awareness and consumer touchpoints.

Relevance Engineering For AI Overviews

Authentic Jobsite Content That AI Understands

Google no longer evaluates an individual website page as a whole but instead looks at fragments and embeddings that align with a specific user intent or expectation.

Mike King makes this clear:

“We’re using embeddings to represent the whole site… and seeing what the distance is for a given page. If anything is too far away, we just delete that content. And every time we’ve done it, the whole site performs better.” – Mike King

Many DataPins users come from old-school SEO practices in which their previous agency engaged in obsolete practices such as mass-producing blog posts and city pages.

We often delete these pages for our clients before they start using our software to realign the embeddings and fragments.

From there, DataPins produces hyper-relevant and semantically dense job reports which are NEVER off-topic or orphaned.

Each job is anchored to authentic customer experiences, reducing the semantic distance between the contractor’s domain and what Google expects to find on that domain.

Relevance Engineering Pin

Software for The SERP Economy, Not the Gig Economy

The emergence of AI Overviews has reduced organic clicks, leading agencies and their clients to look in horror at their Google Analytics 4 reports.

The irony is that DataPins produces far more clicks than traditional SEO strategies because it eliminates practices that violate Google’s Helpful Content Update guidelines, such as mass producing posts and pages.

Still, DataPins is not as singular-minded. It targets outcomes far more impactful than clicks by inserting your brand and business name into Google’s AI synthesis layer.

One thing King relayed was:

“Search has always been a branding channel… you might type in a non-branded query and see branded information and take action… we’ve never taken credit for that.” – Mike King

The loss of organic clicks on GA4 reports may result in some businesses flocking to paid advertising options to replicate concrete traffic projections.

However, that strategy keeps your company mired in a gig economy that relies on conglomerates to drive customers to your local business.

Moving Forward With DataPins for Relevance Engineering

Mike King is 100% right about the SEO industry, which is captured in the following quote:

“What we’re doing is just completely out of phase with what Google is doing… most of our industry does not know what it’s doing right now.” – Mike King

King is referencing the broader SEO community, which is 99.9% of agencies and software developers.

DataPins was engineered to differentiate itself from this archaic and obsolete industry so local contractors could compete in this modern era.

Using structured jobsite pins, DataPins semantically aligns content with Google’s Gemini embeddings and AI Overviews.

The days of “optimizing” pages are long gone, and the new era of relevance engineering is upon us. DataPins was designed for this moment.